Requesty
Together AI Inc. logo
Together AI Inc.

deepseek-ai/DeepSeek-V3

DeepSeek-R1-Distill-Qwen-7B is a 7 billion parameter dense language model distilled from DeepSeek-R1, leveraging reinforcement learning-enhanced reasoning data generated by DeepSeek's larger models. The distillation process transfers advanced reasoning, math, and code capabilities into a smaller, more efficient model architecture based on Qwen2.5-Math-7B. This model demonstrates strong performance across mathematical benchmarks (92.8% pass@1 on MATH-500), coding tasks (Codeforces rating 1189), and general reasoning (49.1% pass@1 on GPQA Diamond), achieving competitive accuracy relative to larger models while maintaining smaller inference costs.

🔧Tool calling
Try deepseek-ai/DeepSeek-V3CompareDocs

Specifications

Context window131K tokens
Max output—
API typechat
AddedMay 30, 2025
Model IDtogether/deepseek-ai/DeepSeek-V3
Data retentionNo
Used for trainingNo
Provider location🇺🇸 US
Privacy policyTogether AI Privacy Policy →

Benchmarks

Released 2024-12
SWE-Bench Verifiedcoding
42.0%

Resolving real GitHub issues from 12 popular Python repositories.

GPQA Diamondreasoning
59.1%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

MMLU Proknowledge
75.9%

Massive Multitask Language Understanding across 57 academic subjects.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality — always test on your own workload.

Pricing

Input / 1M
$1.25
Output / 1M
$1.25
Cache write
—
Cache read
—
Estimated cost
100K input + 10K output$0.14
1M input + 100K output$1.38
10M input + 1M output$13.75

Requesty charges exactly what the upstream provider charges — no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to together/deepseek-ai/DeepSeek-V3.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="together/deepseek-ai/DeepSeek-V3",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Together AI Inc. models

Kimi K2.5
262K ctx$0.50/1M
deepseek-ai/DeepSeek-R1
64K ctx$3.00/1M
SWE-Bench 49%
Qwen/Qwen2.5-72B-Instruct-Turbo
33K ctx$1.20/1M
Qwen/Qwen2.5-7B-Instruct-Turbo
33K ctx$0.30/1M
meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo
131K ctx$0.18/1M
meta-llama/Llama-3.2-3B-Instruct-Turbo
131K ctx$0.06/1M

Frequently asked questions

How much does deepseek-ai/DeepSeek-V3 cost?
deepseek-ai/DeepSeek-V3 is priced at $1.25 per million input tokens and $1.25 per million output tokens when accessed via Requesty. Requesty charges exactly what the upstream provider charges — we don't add markup.
What is the context window of deepseek-ai/DeepSeek-V3?
deepseek-ai/DeepSeek-V3 has a context window of 131K tokens. That's roughly 175 words of input you can fit in a single prompt.
How does deepseek-ai/DeepSeek-V3 perform on benchmarks?
deepseek-ai/DeepSeek-V3 scores 90.2% on MATH, 82.6% on HumanEval, 75.9% on MMLU Pro. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can deepseek-ai/DeepSeek-V3 do?
deepseek-ai/DeepSeek-V3 supports tool calling. You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use deepseek-ai/DeepSeek-V3 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "together/deepseek-ai/DeepSeek-V3". The Quickstart above shows Python, JavaScript and cURL snippets.

Access deepseek-ai/DeepSeek-V3 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll Together AI Inc. models